You are viewing a single comment's thread from:

RE: Introducing Smackdown Kitty

in #steem7 years ago

Because it is not going to target what are most likely new users (under 1000SP) probably both. There is plenty of people who don't self vote their posts, and I remember that @berniesanders barbed me with that in the heated exchange that I initiated with him that led to my being flagged. If those who approve of it form a consensus that it should not hit original posts, then we will follow the will of the people on this. Except the ones who don't want it at all.

Sort:  

I think it is a bit strange to punish behaviour that is the default in the interface. Other than that, no big deal, even though I do upvote my own posts, and I intend to keep doing that. I hardly ever upvote my own comments though, only occasionally to get some comments in the right order or to counteract a flag.

I had been thinking that the default was a problem. I made an issue for it and pull request to change this.

No big deal maybe, but what do you think of the stated issues though? Self appraisal, reward and bandwidth usage?

I'm still thinking about it.

There is a rather large class of people for whom self-upvoting their main posts is the only way to see any return, what with curating being rather slack for many. I also do it for the little bit of extra money and, as a remote second reason, to give myself a pat on the back. Because I can.

I don't really mind self-upvoting, tbh, I mind repetitive self-upvoting of spammy comments. Repetitive, self-upvoted main postings are very visible to everyone, but dito comments are far less easily noticed.

As a solution to self-upvoting spammy comments, I don't think this will work, as the people who do such things are in all probability the same people who will circumvent it by using an extra account or extending the scope of existing circle jerks. That means you can't solve the problem without judging the spamminess of comments and the voting going on. A bot can give an indication something is amiss, but in the end this requires human judgment.

For the purpose of Steemit, I am just a user, and not interested in bandwidth considerations. Get things to work the way you want them, and then fix the bandwidth problems, if any.

Edit: I like experiments, and I would even like to see how this plays out, as long as the bot owners have the good grace to kill the bot when things don't go as intended and the perceived problem isn't solved.

There is a rather large class of people for whom self-upvoting their main posts is the only way to see any return, what with curating being rather slack for many.

This has only been true for a handful of days so the way you've described it here is not really true.

It's precisely the "because I can" angle that the bot is designed to challenge. I get you, it's the incentive and why not? But I feel it's negative over all. We're highlighting an area that requires a systems level change.

In addition, I like experiments so much, that you can use me as a test subject. I will still be upvoting my own main postings, so that should give you something to work with.

Awesome, great attitude you have, disagreeing, and yet willing to try it out. Thank you

As I said in the edit, I would love to see this play out. I also believe the real offenders will circumvent your solution within hours.

This has only been true for a handful of days so the way you've described it here is not really true.

I experience and hear differently.

@l0k1 made the point of saying it was intended for determined offenders, the hope is to add friction to it. You'd be surprised how much UI affects behavior.

You know where I stand, so I am not saying this for myself, but I suspect that if you want more universal support, you should focus on the comments, not the main postings. You run the risk of forcing an opinion on people by bot, including people who aren't doing much wrong.

I'll definitely consider just targeting comments, perhaps you are right. The comments are problematic in a few ways. They worth of them is not really summarized on the post so it doesn't open the commenter to the kind of scrutiny that Steemit relies on to keep abuse in check.

You run the risk of forcing an opinion on people by bot, including people who aren't doing much wrong.

I don't understand what you mean. How is it forcing an opinion?

Also, to say that we should have self upvotes because some people don't want to or can't produce a post worth voting on, doesn't mean we should leave it as it is. Do we make able bodied people use wheelchairs to make wheelies feel better? I think genuinely disabled people find this insulting.

As described in the updated section above, smackdown kitty will be giving upvotes to any user it previously flagged due to self voting. The upvote will only be small, maybe 1%, but it will be another number to the user's comments. Yes, every one of their comments not self-upvoted. And if they resume, the counter continues upwards.

First, I'm not a fan of flagging. I'm especially not a fan of automated flagging. And I'm especially especially not a fan of organizing an online mob to deprive stakeholders of their agency.

Second, the issue is spam, not self-voting, so you're targeting the wrong problem. Votes for low-quality content are harmful to the platform whether they're cast by self-voters, sybil-voters, a group of collaborators, bots, or random strangers. Similarly, votes for high quality content are valuable in all of those cases. If I understand the project, you will cancel out the good with the bad, when you have absolutely no basis to know which dominates, and even if it is currently tilted towards bad, the community hasn't had time to adapt to HF19 yet.

Third, I am particularly offended that @l0k1, who 3 or 4 months ago was writing posts that amounted to "F*&! Steem and its crappy code, I'm going to Dawn Network." now presumes to dictate how I should vote.

Lastly, some of the people doing the self-voting have some serious stakes in their wallets. How long do you suppose it will be 'til they launch their own bot to counter the @smackdown.kitty votes? And that bot has the edge because it will earn curation rewards.

If I could make a suggestion, I would suggest following cheetah's lead and upvoting with very-low steempower so that humans can use that vote as a signal to guide their own voting (and muting) decisions.

I'm aware of your position on bots from our previous discussions, and wouldn't expect that aspect to be attractive to you.

I disagree that there is any agency being removed, rather it is everyone's prerogative on how to use their stake and this extends to down voting / flagging. Mob is just a pejorative term for collective action.

@l0k1 made a good point in a comment earlier

individuals are not qualified judges of their own work

If Steemit should be designed to create the incentive for high quality content to be rewarded, surely self voting should not be so incentivised as it will promote the content of those with stake, rather than the opinion of the populous.

Whether you disagree or agree with the method, what about the question of whether or not self voting is problematic. I think it is, you don't say exactly.

I think you may misremember my position on bots. I'm a big fan of bots for up-votes. My opposition to this is to the arbitrary use of the flag.

By itself, no I don't think self-voting is problematic. It only (potentially) becomes problematic when it is combined with spam (i.e. intentionally low quality content). Even then, I'm not completely sure.

I think the pairing is probably harmful, but do the self-voters hold their earnings? If so, then maybe they're driving up prices and helping the platform more than a high-quality author who dumps half their income every week. (Maybe we should start flagging authors who don't power-up 100%?) Do the self-voters dump their earnings? If so, better voters and better content should win at the long game, which is good for the platform. It might be a case of Bastiat's seen and unseen.

I think you may misremember my position on bots. I'm a big fan of bots for up-votes. My opposition to this is to the arbitrary use of the flag.

My apologies, I did. 😅

The market and economic argument I'm not sure about, but from what I know that is very hard to know. Something that hard to know may not be useful to consider.

By itself, no I don't think self-voting is problematic. It only (potentially) becomes problematic when it is combined with spam (i.e. intentionally low quality content). Even then, I'm not completely sure.

I disagree, I've looked around and there are some good comments with crazy rewards from self votes. They are over valued by the commenter, which is to say that it is not matched by other voters, far far from it.

It depends what we're trying to do here I think. If the authors of posts and comments have that much of a say in valuation I feel that the intended idea of allowing quality stuff to bubble up is not achieved. However if it's to allow those with stake to gain even more interest on their stake then it works well.

I disagree, I've looked around and there are some good comments with crazy rewards from self votes.

True, but there are also self-voted comments that are rewarded appropriately. I go back to my initial post. There's no basis to know which dominates, and the community hasn't had time to adapt to HF19 yet, anyway.

The market and economic argument I'm not sure about, but from what I know that is very hard to know. Something that hard to know may not be useful to consider.

Or maybe humility should suggest restraint of action at scale unless we have some level of confidence that there is a (long term) problem that needs to be solved. And even if there is a problem, maybe there are other solutions involving information exchange among voters that are less confrontational and disruptive than mass flagging.

Gotta run. Thanks for listening to my thoughts.

Well thanks for engaging with it here, it is highly appreciated.

It's clearly an issue, I stand by that. In my opinion it is really against the philosophy of the platform, what I've read of the intentions from the whitepaper, but the proof is in the melting pot of the community. I'm interested to hear other comments too.

Humility is always a good thing but I didn't make myself clear. When such a thing cannot conclusively be known, which the market effects are very likely to be, they are not valuable to consider.

The primary tactics of smackdown kitty are stigmatising, not actually neutralising. She will attempt to do this but unless someone delegates SP to her, she'll just be a little mosquito, leaving red marks everywhere they self vote.

Also, I've seen enough 'anecdotes' of people exploiting self voting, literally powering up, then posting rubbish, and then voting on it, dropping idiotic comments everywhere. I have been watching numerous commentators do the same thing. If the rate of new user signups continues to accelerate how long do you think it is before people who made money scamming and spamming will realise they can just throw 500 bux at this and double it easily in a year? Considering that it is likely to also go up in price at the same time...

Anyway, we'll see by this time next week how it looks. Numbers on the preliminary runs on the bot are showing between 4-15% self upvotes on comments out of the whole of transactions. If that averages about 8%, that's quite a lot. We are also going to gather the info on how much SP these have, and show people the overall demographic, alongside the leaderboard for the most self-voting users.

I have updated the post to include a short summary at the beginning of what and how I think she should smack down this behaviour. Like you are suggesting, perhaps for any user she has previously flagged, if they stop, she instead gives them a 1% upvote, and resets the counter. This way the users will notice that every one of their posts gets an upvote every time they don't self vote.

Self voting is spam, as far as the central purpose of Steem goes - rewarding content that is highly regarded. People only do it because button.

I disagree with this as a blanket statement. Yes, sometimes it is true, but often, it's not. It depends on where the vote is placed. You need to know more about the post/comment to know whether it was spam or not.

For example:

My 15 year old son, @cmp2020, has studied piano for the last 6 or 7 years, and music theory for the last year. When he posts on those topics, his self-vote is far more valuable than many other votes in conveying information to the system. Similarly, at a higher order, @lemouth is a particle physicist and @justtryme90 is a biologist. If they want to promote their own comments to the top of a discussion in their area of expertise, they should certainly be able to.

I'm sure that almost everyone on steemit has specialized knowledge on some topic or other, where their own self-vote should carry more weight than the vote of a random person who happens to see the post/comment.

I think the number of users now has led to a place where there is an audience for every niche. This argument steadily becomes more invalid.

As long as non-experts are able to vote on comments and posts of all topics, the argument will remain valid. You can't assume that all of the people voting are the ones with expertise.

Experts only rule in the world of governments. Here in the wild west, experts are rare, most people are generalists.

Not only that, there's more people who think they are experts than there actually is experts. So the argument still diminishes in validity over time.

I have revised this, we are going to leave the original posts alone, and only mark the comments with very low powered flags, that will slowly escalate linearly with each subsequent self vote by a user.

Sheesh, you're up early 8-).

I think that way of operating will get the most support from the community, as it operates more precisely where the annoyance is.

Can't be your test subject anymore then though, as I don't upvote my comments. I will if you want me to, though.

The kitty is ready, and we have the first draft of the code and today I'll be revising it. Kitty will also generate 'hall of shame' lists of the top 50 self voters every day and I'll be posting it with payout declined. But we are not letting the cat out of the bag until we have tested it on ourselves and whoever wants to - message me (@l0k1) or @personz in the chat to get added to the test list.

Yes, I wake up at about 30 minutes before dawn, every day, my apartmennt has an east facing window and from 7am to about 11am I can sit at my computer, and get direct sun exposure, and yes, I am doing this every day. For my health. I am resetting my body's circadian clocks with this and earthing, to suppress EMF interference with the sensing of the Schumann resonance, which is natures reference clock, and which all living things evolved to depend upon. Biological processes are complex trees of dependent sequences and when they are out of sync, you get sick. It's the number one cause of modern illness.

I agree it would be better to not punish self upvoted posts for the time being, but keep in mind that there are different user interfaces and not all behave the same.

I may add a point about self voting on posts. Since the upvote option is already checked when we format and publish posts, it can be left unchanged mistakenly as well.

Flagging something done in innocence will not be a good option I think. Additionally, even if people self vote on their new posts, how many posts does an average user post? Let's say it's 3 per day (though it will be far less in most cases). Now even if 3 votes are cast on self posts, there are still 7 full votes remaining. Problem occurs when those 7 votes are also given to self via comments.

I started using Steemit on 6 June, 2017 and since then I have commented more than 800 times and posted hardly 40 times. Comments are 20 times more frequent. If I self vote comments, it will be a disaster. The point is, people can do dozens of comments everyday. If they self vote on comments, it is a big problem because (i) it's intentional and (ii) it consumes all their votes and they do not contribute to the community.

My suggestion would be to focus on comment based self voting and may be look out for users who self upvote 100 % of their posts.

Regards,

Ilyas

Fortunately, I, @personz and @the-ego-is-you already discussed this, and between me and @personz, we decided that since the interface defaults to self voting posts, that we would not touch original posts. It would be confusing. For the same reason kitty will not touch under 1000SP accounts, at all. It is only this level and above where you start to see the unsightly large self rewards for a few paragraphs of commentary.

If we succeed in having direct self votes banned in the network consensus, this function and button will be removed.

Yes, you also clearly grasp the disproportionate count of posts versus comments. It is not just an issue of simply too much to look at, it is also so many little targets to pin upvotes to. HF19 rolls around, and that's what people start doing. The comments are a bigger overall problem because they allow users to unroll a lot more of their voting power on their own stuff.

Thanks for the update. I am glad you people are already concerned and looking for the best possible way out.

About the 1000 SP limit, I have a reservation. I have about 164 SP right now and it would take quite some time to reach the 1000 mark.

The folks that I brought to Steemit are still in the single of double figures of SP right now. Now, my point is; there's so much time left for us to reach 1000 SP. Is it ok for us to self vote? It shouldn't be in my opinion because 1-1000 SP is where HABITS are formed and if wrong habits are formed, they will be very difficult to remove.

From a community grooming point of view, I think it would be nice to bring the limit a bit down if possible to include maximum number of people. It would surely increase the work and I am not sure where my suggestion stands when viewed holistically keeping all concerns in consideration. I just wanted to express my opinion though.

Regards,

Ilyas

Self vote away. And remember that the kitten only flags with 1%, and only for 1000+ accounts. We have to wait until tomorrow when we can delegate the needed and necessary SP to her as she cannot operate without around 1000SP!

No, we can't do it to newbies because they don't understand the interface yet, and steemit, inc made the downvote look very menacing. Probably a very large majority of new users barely even know how to use a bitcoin wallet, and this is an intentional design of Steem - that it be approachable to people coming from places like Facebook. Anyone who signed up before The Experiment, which led to HF19, will have a much more blunted emotional response to flags, especially with how it has turned out to be beneficial. And not just, of course, now they can assign themselves substantial rewards, but that they are receiving them in their post rewards as well, even not counting the default upvote.

The default self upvote is 100%, by the way, and I just want to thank you again for helping me realise this. This is also very detrimental to minnows, who lose one of their limited number votes per day to this, before their voting power falls below 20%.

The process of this information propagating to new users is diminished in part due to the fact they are unfamiliar with everything, and there is the issue of the lack of chat integration with the blockchain accounts as well. If there was a trollbox on every page like many exchanges, this information would rapidly infect new users. As it stands, it does not.

This is also why as I progress with my distributed system architecture design and implementation, it is directly the 5th layer of my architecture above the fundamentals, the consensus, the accounts, and cluster membership components, which you can read about here: https://gitlab.com/dawn-network/nexus

In other words, I am making a blockchain type system for user messaging, that ties to an account managemennt system, network consensus and peer discovery process, it will be the 4th element developed, built on the first three, and with this system, it will be possible for all of the Steem interfaces (steemit.com, busy.org, chainbb, and eSteem) to have an integrated, serverless messaging system, which will assist in uniting the community more strongly.

Thank you again for the detailed reply. All my queries are clear. The 100 % default vote is a burden as it drains the voting power. I had discussions about it with some people and I think more people are concerned about this limitation as well.

That's the beauty of it. People own Steemit and show their concern. I love that!

Trollbox is a good idea for steemit.com btw 🤔

Coin Marketplace

STEEM 0.28
TRX 0.22
JST 0.040
BTC 95329.80
ETH 3618.16
SBD 3.87